Title:
学習済みモデルの生成方法、学習済みモデル、表面欠陥検出方法、鋼材の製造方法、合否判定方法、等級判定方法、表面欠陥判定プログラム、合否判定プログラム、判定システム、及び鋼材の製造設備
Document Type and Number:
Japanese Patent JP6973623
Kind Code:
B2
Abstract:
A learned model generation method generates, by using a teacher image that is an image indicating a distribution of a defect portion of a surface of steel and includes a defect map of an equal image size and presence/absence of periodic defects assigned in advance to the relevant defect map, a learned model for which a defect map that is an image indicating a distribution of a defect portion of a surface of steel and having an image size of the equal image size is an input value and a value concerning presence/absence of periodic defects in the relevant defect map is an output value, by machine learning.
Inventors:
Takahiro Koshihara
Hiroaki Ohno
Hiroaki Ohno
Application Number:
JP2020509542A
Publication Date:
December 01, 2021
Filing Date:
October 31, 2019
Export Citation:
Assignee:
jfe Steel Corporation
International Classes:
G01N21/892; B21B38/00; B21C51/00; G06T7/00
Domestic Patent References:
| JP2010185868A |
Foreign References:
| CN108242054A |
Other References:
KOPACZKA et al.,Automated Enhancement and Detection of Stripe Defects in Large Circular Weft Knitted Fabrics,2016 IEEE 21st International Conference on Emerging Technologies and Factory Automation (ETFA),2016年
Attorney, Agent or Firm:
Sakai International Patent Office